Objective To compare the need for positive pressure ventilation (PPV) by bag and mask and by bag and endotracheal tube in newly born term infants with vertex presentation delivered by non-urgent caesarean section under regional anaesthesia or non-instrumental vaginal delivery.
Design Cross-sectional study.
Setting 35 public hospitals in 20 Brazilian state capitals.
Patients 6929 inborn infants without congenital anomalies, with gestational ages from 370/7 to 416/7 weeks with vertex presentation, born between 1 and 30 September 2003.
Intervention Non-urgent caesarean versus non-instrumental vaginal delivery. Non-urgent caesarean was defined as delivery occurring in the absence of prolapsed cord, third trimester haemorrhage, failure of labour induction, fetal distress or non-clear amniotic fluid.
Main outcome measures PPV with bag and mask and with bag and endotracheal tube. Both outcomes were adjusted for potential confounding variables by logistic regression analysis.
Results 2087 infants were born by non-urgent caesarean and 4842 by non-instrumental vaginal delivery. Non-urgent caesarean delivery under regional anaesthesia compared to vaginal delivery under local or no anaesthesia increased the risk of bag and mask ventilation (OR 1.42, 95% CI 1.07 to 1.89) adjusted for number of gestations, maternal hypertension and birth weight. Ventilation with bag and endotracheal tube was associated only with low birth weight, adjusted for delivery mode and twin gestation.
Conclusions Term neonates with vertex presentation and clear amniotic fluid born by non-urgent caesarean section under regional anaesthesia need to be assisted at birth by health professionals skilled in PPV.
Statistics from Altmetric.com
The World Health Organization advocates an ‘optimal’ national caesarean delivery rate of between 5% and 15% of all births.1 Caesarean delivery is linked to wealth at national and individual level.2 The low rates of caesarean delivery in poor countries represent a failure to care for the most vulnerable mothers. On the other hand, in the more advanced regions of the world, caesarean delivery rates are rising exponentially.2 Europe, USA, South America and Eastern Asia have reported rates in excess of 30–40%.3,–,7 In Brazil, in 2007, there were 2 640 897 live births with a gestational age of 37–41 weeks, and 46% of them were born by caesarean section. In the most developed areas of the country, such as the south and southeast regions, this rate reaches 53% of term infants.8
What is already known on this topic
▶. Population-based studies report neonatal vitality at birth after elective caesarean section and also use Apgar score at 5 min as proxy for ‘birth depression’.
▶. A low Apgar score alone is not a specific indicator for intrapartum compromise and, besides, resuscitative interventions modify the components of the Apgar score.
What this study adds
▶. For term infants with vertex presentation and clear amniotic fluid, non-urgent caesarean delivery under regional anaesthesia is a risk factor for ventilation with bag and mask.
▶. Term neonates with vertex presentation and clear amniotic fluid born by non-urgent caesarean section under regional anaesthesia need to be assisted at birth with positive pressure ventilation.
Term neonates born by elective caesarean delivery have an increased risk of respiratory distress syndrome, transient tachypnea of the newborn, persistent pulmonary hypertension and need for supplemental oxygen and mechanical ventilation.9 10 In addition to respiratory morbidity, infants delivered by elective caesarean section have poorer outcomes as regards breastfeeding and are more likely to require special care admission, compared to those born by vaginal delivery.11 In term, singleton vertex births, caesarean deliveries with no labour complications or procedures have a 69% higher risk of neonatal mortality than planned vaginal deliveries,12 but consideration should be given to fetal demise in ongoing pregnancies.13
Some population-based studies have reported neonatal vitality at birth after elective caesarean section and used Apgar score at 5 min as proxy for ‘birth depression’ since this information is available from perinatal databases.11 14 15 However, a low Apgar score alone is not a specific indicator for intrapartum compromise. Resuscitative interventions, generally started before 1 min of life, modify the components of the Apgar score.16 Few studies address resuscitative procedures at birth. Annibale et al17 evaluated 11 702 term gestations in two American hospitals and found a higher incidence of bag and mask ventilation and intubation in all caesarean sections compared to vaginal deliveries, but indications for caesarean section, the presence of fetal distress and the type of anaesthesia were not specified. Parsons et al18 and Atherton et al19 compared resuscitative procedures between vaginal and caesarean deliveries for term, singleton gestations in Australia, using population-based studies. In a subgroup analysis of non-instrumental spontaneous vaginal deliveries versus elective repeat caesarean sections under regional anaesthesia without fetal distress, the authors did not find an increase in the need for neonatal tracheal intubation at birth. Indications for deliveries, as stated by the authors, were not prospectively assessed, neither was the delivery room neonatal care uniform. Gordon et al,20 in a hospital-based cohort prospective study, evaluated 44 938 singleton term gestations. Comparison of elective caesarean section under regional anaesthesia with spontaneous vaginal delivery showed that 7.8% and 4.2%, respectively, of the newborn infants needed bag and mask ventilation in the delivery room, but the need for advanced resuscitative procedures (intubation and/or circulatory support) was not different between the groups. Finally, Kamath et al10 retrospectively evaluated 672 women with one prior caesarean delivery and a singleton term pregnancy grouped by the intention to deliver by caesarean or vaginally. More neonates in the intended vaginal group required bag and mask ventilation (3.3%) and endotracheal intubation (2.4%) in the delivery room than neonates born in the intended caesarean group (2.3% and 0.6%, respectively).
In face of the increasing rates of surgical deliveries worldwide and the debate regarding human and material resources needed for neonatal resuscitation following such births, the aim of this study was to examine whether delivery by non-urgent caesarean section compared to the vaginal route increases the need for positive pressure ventilation (PPV) by bag and mask and by bag and endotracheal tube at birth in newborn infants with a gestational age between 370/7 and 416/7 weeks.
This cross-sectional study enrolled term infants born at 35 hospitals in 20 Brazilian state capitals between 1 and 30 September 2003. The hospitals were selected in each capital according to the following criteria: (1) they had to be public maternity clinics with the greatest number of births in the city and with more than 90% of patients financed by the public system; (2) they had to be situated in capital cities with at least 5000 live births in the year 2002.8 The infants selected represented 4–8% of the liveborn infants in each state. The institutional research ethics committee associated with the main investigators and the clinical board of each participating institution approved the research project.
At each hospital, one paediatrician collected data on the characteristics of the institution, the material/equipment available in the delivery room of each neonatal resuscitation unit21 and the human resources available for delivery room neonatal care according to international guidelines.22 Maternal and neonatal data were collected daily for each infant born alive during the study period. All health professionals in charge of neonatal care in the delivery room and involved in data collection were unaware of the specific goals of this study.
Inborn infants were included if they had vertex presentation, had a gestational age between 370/7 and 416/7 weeks as defined by the best obstetric estimate, and were born by caesarean section or non-instrumental vaginal delivery, with clear amniotic fluid at birth and without congenital malformations. For all caesarean deliveries, the paediatrician noted which of the following indications were present: previous caesarean section, failure of labour induction or progression, cephalopelvic disproportion, fetal distress, non-vertex presentation, placenta abruption, placenta previa, prolapsed cord or other causes. Non-urgent caesarean section was defined when the infant was delivered in the absence of: prolapsed cord, third trimester haemorrhages, fetal distress, or failure of labour induction or progress. Deliveries performed under general anaesthesia were also excluded.
Recorded information from the delivery room included the number and category of health professionals who assisted the newborn as well as the resuscitative procedures carried out according to guidelines established by the American Academy of Pediatrics and the Brazilian Society of Pediatrics in 2000.22 Neonatal resuscitation interventions were categorised as: PPV only with self-inflating bag and mask for at least 30 s; ventilation with self-inflating bag and endotracheal tube for at least 30 s; PPV with chest compressions and/or medication. During this period, ventilation was applied with 100% oxygen.
Sample size was defined as a minimum of 1000 term infants in each group defined by delivery route, considering a baseline need for PPV with bag and mask of 3% for vaginal delivery versus 6% for caesarean section, with an α error of 1% and a power of 90%.
Comparison of maternal and neonatal characteristics according to mode of delivery was carried out using χ2 or Fisher's exact test and by t test or Mann–Whitney test. Perinatal factors associated with the need for bag and mask ventilation and those associated with endotracheal intubation were analysed by stepwise logistic regression. As independent variables, maternal and neonatal characteristics were analysed and first included in the model if univariate analysis showed p<0.20. All analyses were carried out using SPSS 16.0.
In the study period, 11 922 children were born alive in the 35 maternity hospitals. Among these, 9797 infants had a gestational age between 370/7 and 416/7 weeks and were without congenital malformations, while 2801 were excluded due to the following conditions: non-vertex presentation (347), instrumental vaginal delivery (335), prolapsed cord (14), abruption placenta (44), placenta previa (22), failure of labour induction or progress (472), fetal distress (495), non-clear amniotic fluid (1042), delivery under general anaesthesia or unknown anaesthesia type (30). The indication for caesarean section was not noted in 51 cases, which were also excluded. Therefore 6945 infants were enrolled in the study, 4852 of whom were born by non-instrumental vaginal delivery and 2093 (30%) by non-urgent caesarean section.
Of the 35 institutions, 23 (66%) were Baby-Friendly hospitals, 31 (89%) were reference centres for high-risk pregnancies and 26 (74%) were teaching hospitals for medical residents in paediatrics. These 35 hospitals had 118 neonatal resuscitation beds, with an average of three (95% CI 3 to 4) in each obstetric centre. The equipment available for resuscitation for these 118 beds was: radiant heat source, oxygen and vacuum in 100%, self-inflating bags and masks in 98% and laryngoscopes, endotracheal tubes and medications in 95%. On average, one or two paediatricians, three registered nurses and five nurse technicians worked per shift in the delivery rooms of each hospital. Of the 874 paediatricians in the 35 hospitals, 94% had received 8 h of theoretical and practical training in neonatal resuscitation.
Maternal and neonatal characteristics according to mode of delivery are displayed in tables 1 and 2. The frequency of non-adolescent mothers, twin gestation, hypertension or diabetes during pregnancy, prolonged rupture of membranes and number of prenatal visits was higher among non-urgent caesarean section deliveries. Anaesthesia varied according to delivery route: 94% of mothers delivering vaginally had local or no anaesthesia, while 100% of those undergoing caesarean section had regional anaesthesia. Neonates born by caesarean section were 100 g heavier than those born by vaginal delivery. Regarding resuscitation procedures (table 2), 158 (3.3%) infants born by vaginal delivery received bag and mask ventilation, 13 (0.3%) bag and endotracheal tube ventilation and four (0.1%) PPV with chest compressions or medication at birth, while 98 (4.7%) born by caesarean section received bag and mask ventilation, eight (0.4%) bag and endotracheal tube ventilation and six (0.3%) PPV with chest compressions or medication at birth. Therefore, the need for PPV with bag and mask was 1.46 (95% CI 1.13 to 1.89) higher in neonates born by non-urgent caesarean section compared to vaginal delivery. For PPV with bag and endotracheal tube, the OR was 1.43 (95% CI 0.54 to 3.68).
Univariate analysis of variables associated with PPV by bag and mask or with PPV by bag and endotracheal tube are shown in table 3. Upon multivariate analysis, the following perinatal factors were associated with the use of only bag and mask ventilation in this population (table 4): first gestation, maternal hypertension, caesarean delivery in association with anaesthesia type, and birth weight less than 2500 g (Hosmer and Lemeshow test, p=0.671). Non-urgent caesarean section under regional anaesthesia compared to vaginal delivery performed without or with local anaesthesia increased the need for ventilation with bag and mask (OR 1.42; 95% CI 1.0 to 1.89) adjusted for other variables.
Multivariate analysis showed that only low birth weight, adjusted for delivery mode and twin gestation, was associated with bag and endotracheal tube ventilation in this population (Hosmer and Lemeshow test, p=0.493) (table 4).
This study shows that, for term infants with vertex presentation and clear amniotic fluid, non-urgent caesarean section under regional anaesthesia is a risk factor for PPV with bag and mask at birth compared to non-instrumental vaginal delivery, when these outcomes are adjusted for confounding variables.
Definitions of elective caesarean delivery vary considerably among studies.23 The definition of non-urgent caesarean section adopted here and the decision to exclude non-vertex presentations, any meconium or blood stained amniotic fluid, instrumental vaginal deliveries and caesarean sections performed under general anaesthesia, allowed evaluation of the subset of deliveries that did not have a clear maternal and/or fetal risk for birth asphyxia and need for neonatal resuscitation at birth.24 Indications of caesarean section in all infants were prospectively determined by each local investigator at birth, blind to the primary goals of this study. This approach differs from that found in literature which excludes maternal and neonatal conditions related to medically indicated surgical delivery based on information retrieved from birth certificates and/or hospital discharge databases.17,–,20 23 25 On the other hand, the definition of non-urgent caesarean section does not equal no-indicated risk (NIR) for caesarean delivery in mothers who have a full-term, singleton, vertex presentation birth, without medical risk factors or labour and delivery complications, and no prior caesarean.26 NIR or elective caesarean delivery may be scheduled to accommodate patient or physician convenience.27 In fact, in a cross-sectional study performed in a public (2889 deliveries/year, with a caesarean section rate of 19%) and a private maternity hospital (2911 deliveries/year, with a caesarean section rate of 83%), in São Paulo, Brazil, in the private hospital non-medical factors were more highly associated with caesarean section than biological or clinical factors related to pregnancy.28
Other studies regarding the impact of caesarean delivery on neonatal vitality at birth report conflicting results. Liston et al11 studied 142 929 term deliveries from 1998 to 2002 in Nova Scotia, Canada, retrieving their data from the Nova Scotia Atlee Perinatal Database. Depression at birth was defined as a delay in initiating and maintaining respirations or a 5 min Apgar score lower than 3. Interestingly, caesarean delivery in labour was more likely to be associated with increased risks of depression at birth compared to spontaneous and assisted vaginal delivery than caesarean delivery without labour: these findings may be related to the indications for a caesarean section as opposed to the method of delivery itself.
De Luca et al14 evaluated 56 549 late preterm and term deliveries from 1982 to 2004 in a single centre in Geneva, Switzerland. Depression at birth was defined as an arterial cord blood pH <7.10 and/or a venous cord blood pH <7.15 and/or a 5 min Apgar score <7. For late preterm infants, depression at birth was more frequent after caesarean section before the onset of labour and before rupture of membranes (41%) compared to elective vaginal delivery (30%), but for term infants the mode of delivery did not change this outcome.
Kilsztajn et al15 studied 495 745 singleton liveborn infants in São Paulo State, Brazil, in 2003, with a birth weight between 1501 and 3999 g, a gestational age between 32 and 41 weeks and no congenital malformations. The caesarean section rate in this cohort was 51% and poor vitality at birth was defined as a 5 min Apgar score <7. Adjusted for obstetric, demographic and social factors, mode of delivery was not associated with low Apgar score.
All these studies differed from ours in terms of types of deliveries included, collection of information from databases not designed ‘a priori’ to provide answers to specific questions, and different definitions of birth depression. Also, it is noteworthy that all definitions used for birth depression are dependent on resuscitative measures applied in the delivery room, which are not described. The present investigation specifically collected data regarding the indication for caesarean section and all neonates were assisted in the delivery room by paediatricians trained in resuscitation by the Neonatal Resuscitation Program of the Brazilian Pediatric Society. This program offers a structured 8 h theoretical and practical training course that covers basic information on initial steps in resuscitation, ventilation with bag and mask, chest compressions, endotracheal intubation and medication.29
The study by Kamath et al10 was one of the few to address resuscitative procedures at birth. The authors showed an increased need for bag and mask ventilation and endotracheal intubation in infants born by elective vaginal delivery compared to elective caesarean delivery in a group of mothers with a previous caesarean section. As the authors discuss, neonates born after failure of elective vaginal delivery may require more resuscitation due to fetal distress. The striking differences in enrolment criteria compared with our study preclude comparisons.
The results in the present study can be compared to those presented by Gordon et al20 and Atherton et al,19 both of whom enrolled term singleton gestations and evaluated a subset of patients born by elective caesarean section under regional anaesthesia. In these studies, the frequency of bag and mask ventilation varied between 5% and 8% for the caesarean section newborns and 4% for the vaginally delivered infants and intubation was needed, respectively, in 0.08–0.3% and 0.06–0.2% of infants. These data show that the need for bag and mask ventilation is increased in vertex infants born by non-urgent caesarean section under regional anaesthesia. All studies, including ours, show a slight increase in the need for ventilation by bag and tracheal tube in the delivery room, but they do not have enough power to provide a definite result. Even though the populations enrolled in the various reports are large, among low risk groups the number of intubated infants is only 14/23 090 live births in Atherton et al,19 59/24 666 in Gordon et al,20 and 21 out of 6945 newborn infants in our study. Therefore, only a meta-analysis of the available data would have enough power for a conclusion on the need for intubation after non-urgent caesarean section.
These findings should be analysed in the broader context of the long-term outcome of infants who require resuscitative interventions at birth. The Avon Longitudinal Study of Parents and Children cohort shows that infants who are resuscitated at birth have an increased risk of a low IQ score at 8 years of age even if they remain healthy during the neonatal period.30 Therefore, it is a matter of concern that the rising rates of non-medically indicated caesarean deliveries in the developing and developed world may result in increasing rates of neonatal resuscitative interventions, as shown in this study, which may contribute to long-term developmental difficulties in these infants.
The results of the present study indicate that term neonates with vertex presentation and clear amniotic fluid born by non-urgent caesarean delivery under regional anaesthesia need to be assisted at birth by health professionals skilled in PPV.
The authors thank all Brazilian Neonatal Resuscitation Program Local Investigators who helped with data collection, the Brazilian Society of Pediatrics Foundation who supported the study, and Adriana Sanudo for statistical analysis.
Funding The Brazilian Society of Pediatrics Foundation supported this study.
Competing interests None.
Ethics approval The institutional review boards of the Federal University of São Paulo and the clinical board of each participating institution approved the study protocol.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.